# Constraints residues in different chain in protein-protein docking

3 posts / 0 new
Constraints residues in different chain in protein-protein docking
#1

Hello,

I want to dock two proteins. I have some constraints between residues of different chain (here chain A and B).

For eaxmple:

For residues number 15 of chain A, and residue number 271 of chain B, I have the following information:

The minimum distance between residues is 0, and the maximum distance is 6.

How can I add constrain for something like this case (I mean when residues are in different chain)? How can I specify that residues are in diffrent chain?

Thank you very much

Category:
Post Situation:
Sat, 2020-11-28 10:24
nasim.soltani58

Hi!

The documentation for the constraints system is here. It takes a bit to digest but is very useful.

For your case I would look to use the AtomPair constraint. It will allow for definition of a constraint between an atom on residue 15 of chain A and an atom on residue 217 of chain B. For your desired distances a FlatHarmonic function is relatively simple. Finally, You'll want to wrap AtomPair constraints for each pair of atoms (or each logical pair) in an AmbiguousConstraint.

An example of a few atom pairs would look like this:

AmbiguousConstraint
AtomPair CA 15A CA 217B FLAT_HARMONIC 0 0.5 6
AtomPair CA 15A CB 217B FLAT_HARMONIC 0 0.5 6
.
.
.
AtomPair N 15A O 217B FLAT_HARMONIC 0 0.5 6
END

Many constraint types recognize residue definitions using the "pdb numbering" format where the chain identifier follows the residue number, as in the example. Many can also utilize "pose numbering" where the system is numbered from 1 to N as in the input file. In the latter case, if both chains start at 1 and end at 300, then your constraint type definition would use 15 and 517 as the residue numbers.

The flat harmonic function scores as 0 for any distance between the two atoms from 0 to 6 angstroms. Deviations above 6 will incur a harmonic penalty with a standard deviation of 0.5.

In the above, you only need to generate the constraint once for each pair. Adding additional constraints from CA 217B to CA 15A would double the penalty if outside 6 angstroms.

Don't forget to add the atom_pair_constraint term to your score function using -score::set_weights atom_pair_constraint 1.0

Hope that helps!

Sun, 2020-11-29 10:04
nannemdp

Thank you very much for your help.

Mon, 2020-11-30 12:24
nasim.soltani58